Validity of the Instruments

According to Fornell and Larcker (1981), there are two criteria to evaluate the convergent validity of the measurement scales. First, all the factor loadings of the indicators should be significant and exceed 0.70. Second, the average variance extracted (AVE) of each construct should exceed the variance caused by measurement errors in that dimension (i.e. AVE should exceed 0.5). In order to test the validity of the three major constructs (i.e. retail service quality, product quality, and customer loyalty) and their corresponding dimensions included in this study, factor analysis and the calculation of AVE are conducted.

A factor analysis is applied to evaluate the constructs by including relevant variables and excluding irrelevant variables so as to validate the module. In this research, 200 datasets were obtained from the customer survey and used. According to Kaiser (1974) and Janssens et al. (2008), a factor loading is statistically significant if it is equal to or greater than 0.4 for a sample size of 200 (see Table 4.1). Furthermore, Kaiser (1974) advocated that variables with cross loading are insignificant and require deletions. Therefore, variables with a factor loading lower than 0.4 and/or with cross loading had been removed in this study.

Table 4.1 Significant factor loadings for different sample sizes

Factor loading

Sample size necessary to be significant (at the level of 0.50)

0.3 350 0.35 250 0.4 200 0.45 150 0.5 120 0.55 100 0.6 85 0.65 70 0.7 60 0.75 50

The results of factor loading of all variables under the three constructs, i.e. retail service quality, product quality, and customer loyalty, are shown in Table 4.2, Table 4.3 and Table 4.4 respectively. Although the factor loadings of all variables are greater than 0.4 (Kaiser, 1974; Janssens et al., 2008), cross loading is seen in some variables, thus have to be deleted. Regarding retail service quality (see Table 4.2), cross loading is seen in three variables - “Hunter Douglas handles your complaints immediately”, “Hunter Douglas handles your complaints directly”, and “Hunter Douglas shows sincere interest in solving your problems” - and therefore these three variables have been deleted. The resulting variables included in the construct of retail service quality are updated and shown in Table 4.5. In

regard to the variables under the construct of product quality, as no variables have a factor loading of less than 0.4 and/or cross loading, all variables remained (see Table 4.3). Concerning customer loyalty (see Table 4.4), the two variables – “You do not favour other department stores that offer curtains with more attractive prices” and “You do switch to other department stores when the quality of Hunter Douglas presently perceived discontinues” - were deleted due to cross loading. The resulting variables included in the construct of customer loyalty are shown in Table 4.6.

Table 4.2 The factor loading of all variables regarding retail service quality

Factor

Variables 1 2 3 4 5

Staff are courteous to customers .677 .196 .171 .113 .195 Staff provide prompt quotations .692 .189 .237 .207 .067 Staff provide accurate quotations .782 .135 .140 .104 .233 Staff can be easily approached and consulted .843 .106 .143 .022 .163 Staff provide accurate information about Hunter Douglas

(e.g. history, opening hours and product information) .678 .065 .119 .264 .042 Staff give personal attention .649 .217 .113 .360 .136 You find the transactions with Hunter Douglas are safe .207 .838 .194 .146 .059 Hunter Douglas accepts most major payment methods .187 .724 .369 .093 .134 Operating hours of Hunter Douglas are convenient .234 .739 .057 .177 .168 There are visually appealing physical facilities (such as

display, shop decoration and tidiness) in Hunter Douglas .199 .129 .752 .039 .094 There are visually appealing service materials (such as

collection book and catalogue) in Hunter Douglas .096 .284 .744 .065 .095 Staff of Hunter Douglas are dressed neatly and tidily .204 .091 .752 .208 .172 Store layout makes it easy for you to find what you need .197 .074 .747 .217 .181 The service promises (such as return, 3 years of

warranty, and product quality) enhance the reliability of Hunter Douglas

.278 .211 .136 .807 .169 When Hunter Douglas promises to do something by

certain time, it will do so (such as the delivery and installation time)

.322 .190 .149 .769 .164 Hunter Douglas handles your complaints immediately .163 .081 .261 .551 .503

Hunter Douglas handles your complaints directly .202 .056 .245 .483 .650

Hunter Douglas shows sincere interests in solving your

problems .173 .109 .263 .509 .592

The hotlines of Hunter Douglas are seldom busy .165 .173 .123 .073 .871

You are satisfied with leaving message through the

maintenance hotline .163 .060 .127 .174 .888

Hunter Douglas quickly responds to your enquiries or complaints after you left voice messages via the maintenance hotline

.137 .099 .066 .050 .909

Table 4.3 The factor loading of all variables regarding product quality

Factor

Variables 1 2 3

Hunter Douglas provides diverse types of blinds for your selection

.795 .196 .343

Hunter Douglas collections have a range of differentiated colours to cater for your needs

.858 .127 .281

Hunter Douglas collections have a range of differentiated materials to cater for your needs

.888 .080 .023

You are satisfied with the functions of blinds provided by Hunter Douglas (such as dust-resistance and UV-resistance)

.754 .323 .046

The blinds provided by Hunter Douglas have an appealing appearance

.203 .917 .170 The blinds provided by Hunter Douglas are stylish .130 .874 .202 The blinds provided by Hunter Douglas are innovative .222 .924 .170 Hunter Douglas provides products of high quality .272 .379 .752 The blinds provided by Hunter Douglas have high durability .128 .130 .903

Table 4.4 The factor loading of all variables regarding customer loyalty

Factor

Variables 1 2

You share positive opinions about Hunter Douglas with other people .700 .378 You recommend Hunter Douglas to someone who seeks advice in curtains .727 .385 You encourage relatives and friends to shop at Hunter Douglas .801 .228 You consider Hunter Douglas as the first choice when buying curtain next

few years .807 .261

You do not favour other department stores that offer curtains with more

attractive prices .770 .400

Searching and switching to another new retail store is just not worth the

time and effort .790 .070

You do switch to other department stores when the quality presently

perceived at Hunter Douglas discontinues .432 .414

You want to build a good relationship with Hunter Douglas .685 -.114

You are loyal to a particular staff .276 .760

You have no intention to shop in the future because of a particular

employee -.024 .889

Table 4.5 Resulting variables regarding retail service quality after the factor analysis

Factor

Variables 1 2 3 4 5

Staff are courteous to customers .688 .203 .172 .079 .176 Staff provide prompt quotations .713 .208 .235 .140 .033 Staff provide accurate quotations .782 .134 .145 .101 .218 Staff can be easily approached and consulted .829 .095 .145 .051 .156 Staff provide accurate information about Hunter Douglas

(e.g. history, opening hours and product information) .700 .085 .122 .201 .010 Staff give personal attention .662 .222 .127 .335 .126 You find the transactions with Hunter Douglas are safe .202 .834 .194 .157 .052 Hunter Douglas accepts most major payment methods .211 .747 .360 .020 .098 Operating hours of Hunter Douglas are convenient .236 .740 .060 .176 .161 There are visually appealing physical facilities (such as

display, shop decoration and tidiness) in Hunter Douglas .191 .125 .757 .040 .084 There are visually appealing service materials(such as

collection book and catalogue) in Hunter Douglas .097 .288 .746 .043 .076 Staff of Hunter Douglas are dressed neatly and tidily .211 .095 .766 .182 .151 The store layout makes it easy for you to find what you

need .208 .082 .761 .181 .156

The service promises (such as return, 3 years of warranty, and product quality) enhance the reliability of Hunter Douglas

.271 .188 .185 .863 .192 When Hunter Douglas promises to do something by

certain time, it will do so (such as the delivery and installation time)

.305 .158 .199 .848 .197 The hotlines of Hunter Douglas are seldom busy .169 .163 .156 .098 .887

You are satisfied with leaving message through the

maintenance hotline .170 .055 .166 .191 .893

Hunter Douglas quickly responds to your enquiries or complaints after you left voice messages via the maintenance hotline

.135 .084 .102 .093 .930

Table 4.6 Resulting variables regarding customer loyalty after factor analysis

Factor

Variables 1 2

You share positive opinions about Hunter Douglas with other people .715 .354 You recommend Hunter Douglas to someone who seeks advice in curtains .750 .380 You encourage relatives and friends to shop at Hunter Douglas .806 .215 You consider Hunter Douglas as the first choice when buying curtain next few

years .809 .243

Searching and switching to another new retail store is just not worth the time and

effort .791 .052

You want to build a good relationship with Hunter Douglas .689 -.106

You are loyal to a particular employee .306 .788

The validity of the factor analysis was tested using the Kaiser-Meyer-Olkin (KMO) measure of sampling adequacy test and the Bartlett’s test of Sphericity. The results of these two tests are exhibited in Table 4.7. First, applying the KMO measure of sampling adequacy test to retail service quality, product quality and customer loyalty resulted in values of 0.858, 0.819, and 0.878 respectively, which were greater than the recommended value of 0.7 (Kaiser, 1970; 1974) and indicated a meritorious degree of common variance (see Table 4.8). Second, the values (p = 0.000) of the Bartlett’s test of Sphericity for the three constructs are significant, thereby proving the factor analysis appropriate and valid. These results jointly indicated that the structural model has satisfactory fit and is meaningful.

Table 4.7 Results of KMO and Bartlett's test on retail service quality

Kaiser-Meyer-Olkin (KMO) Measure of Sampling

Adequacy Test

Sig. of Bartlett's Test of Sphericity

Retail service quality 0.858 0.000

Product quality 0.819 0.000

Customer loyalty 0.878 0.000

Table 4.8 An interpretation of the KMO measure (Janssens, 2008)

KMO Value Degree of Common Variance

0.90 to 1.00 Marvellous 0.80 to 0.89 Meritorious 0.70 to 0.79 Middling 0.60 to 0.69 Mediocre 0.50 to 0.59 Miserable 0.00 to 0.49 Don't Factor

Apart from factor analysis, the calculation of AVE is also required to test the validity of the instrument. The correlation matrices of the datasets obtained from the customer survey and the results of the calculation of the AVEs are shown in Table 4.9, Table 4.10 and Table 4.11. The correlation matrices were used to examine all potentially overlapping constructs and to calculate the AVEs of the

corresponding dimensions. As shown in the tables, the AVEs of the dimensions all exceed 0.5, showing convergent validity for all the dimensions under the three constructs according to Fornell and Larcker (1981). Furthermore, the construct is confirmed valid if the AVE is greater than the squared correlation between the construct and other constructs (Fornell and Larcker, 1981). With reference to Table 4.9, Table 4.10 and Table 4.11, as all of the diagonal elements, which are the square root of the AVEs of the corresponding dimensions, are higher than the correlation coefficient between the target dimensions, the discriminant validity of all the dimensions in this study is assured. As a result, the three constructs are proved to be valid and in other words, the three constructs are relevant and there are no irrelevant constructs.

Table 4.9 The correlation matrix and resulted AVE regarding retail service quality

AVE 1 2 3 4 5 Personal Interaction 0.535 0.731 - - - - Policy 0.600 0.510** 0.775 - - - Physical Aspects 0.619 0.471** 0.484** 0.787 - - Reliability 0.732 0.549** 0.430** 0.425** 0.856 - Problem Solving 0.820 0.379** 0.310** 0.329** 0.406** 0.906

** The correlation is significant at the level of 0.01.

Table 4.10 The correlation matrix and resulted AVE regarding product quality

AVE 1 2 3

Product Features 0.681 0.825 - -

Aesthetics 0.820 0.442** 0.672 -

Customer-perceived Quality 0.690 0.451** 0.483** 0.831

** The correlation is significant at the level of 0.01.

Table 4.11 The correlation matrix and resulted AVE regarding customer loyalty

AVE 1 2

Customer Loyalty to the Store 0.580 0.762 -

Customer Loyalty to Employees 0.707 0.376** 0.841

In document Exploring customer relationship management using simulation modelling in the retail sector (Page 82-90)